The most noted magical formula of antiquity was what

Clement of Alexandria referred to as "the so-called Ephesian

letters, famous among many."1 The great mass of magical
books which have been preserved have come from another
region. But there can be little doubt that the magical rolls
which the first Christians of Ephesus burned to the value of
fifty thousand pieces of silver 2 were but a small part of those
in use. If the Ephesian climate were as propitious as the
Egyptian, we should doubtless have numerous documents of
magic from the Asian capital. In any case the eeo&La ypa4p.ar7a
constituted the magical formula par excellence in the Hellenistic
world.
By the middle of the fourth century before Christ they were
already famous. Anaxilas Comicus speaks of "auspicious
Ephesian letters," Menander of "evil-averting Ephesian
spells."3 A Cretan lead tablet, quoted below, shows them
actually in use at this date. They were known to Plutarch,
Marcus Aurelius, Clement of Alexandria, and the various
lexicographers and paroemiographers.4 Theie can be little
doubt that an allusion to the famous formula is hidden in a
corrupt passage of Apuleius.5 The magic papyri of the third
and fourth centuries repeat three or four of the six words and,
as I shall try to show, the Testament of Solomon, a Christian
writing of the same era, seems to mention two.
1Strom. v, 8, 45.2 (p. 242 Sylb., 672 Pott.).
2 Acts, 19: 19.
3 Meineke, Frag. Com. Gr. III (Berlin, 1840), 345; Iv (1841), 181; Kock,
Corn. Att. Frag. II, 1 (Leipzig, 1884), 268; II, 2 (1888), 108.
4 See Kuhnert in Pauly-Wissowa, Hild in Daremberg-Saglio, and footnotes
below for the various references.
5 Met. xI, 17 (789); cf. L. Stengel, Rh. Mus. xvI (1861), 34 f.; Dieterich,
Mithrasliturgie (Leipzig, 1903), 38.
128 [1923
Vol. liv] The Ephesia Grammata 129
The "original, genuine" Ephesia grammata were six in
number, and on the authority of Androcydes the Pythagorean,
are given by Hesychius and Clement of Alexandria, who alone
dare to name them, as acrKLov, KaraaKLov, X\i, rcrpac, baVaIva,uvevs,
and a'aLa (aiatov, Hesych.).6 Clement informs us that, like
dactylic hexameter, they were invented by the Idaean Dactyli.7
According to Pausanias, the Atticistic lexicographer, they
"seem to have been written indistinctly and obscurely on the
feet, girdle, and crown of Artemis," 8 a sentence which reveals
principally the author's uncertainty. From the same tradition
come perhaps the lines of the great Paris magic papyrus
which name certain permutations of ba,uvaw as the 'ypaCLuuara
incised on the scepter of Artemis-Selene.9
As to the origin and meaning of the mysterious formula,
ancients and moderns have made their guesses, all alike unconvincing.
10 We can well agree with the lexicographers and
paroemiographers that they were aavbera.11 The form of the
words and the literary tradition as to their origin have suggested
that they came from Phrygian or Cretan sources and
were taken over by the worship that developed about Artemis
6 Hesychius, Lexicon, s.v.; Clement, I.c.
7 Strom. I, 15, 73.1 (p. 132 Sylb., 360 Pott.). Roscher's attempted rearrangement
of them so as to form a hexameter is artificial and contradicts the
evidence (Philol. LX [1901], 88 f., accepted by Gruppe, Griechische Mythologie,
II [Minchen, 1906], 884); see also below, p. 132, note 23.
8 As quoted by Eustathius, ad Homer. Od. T 247. The remark of Arsenius,
'Iowta, xxv, 22, b7rl roi 7roal rijs ieeaias 'AprT$u6ors a 'yeypauPUe'avaav eTra, is
possibly no more than a well-intentioned attempt to simplify the involved
statement of Pausanias, but may describe an image he had seen.
9 Lines 2844-2848, bapvwo, at.vouJoevecbaa,/ acaaavppab, a,Ivo6ba/ia; Wessely,
"Griechische Zauberpapyrus von Paris und London," in Denkschriften der
kaiserl. Akad. d. Wissenschaften, Phil.-hist. Classe, xxxvi, 2 (Vienna, 1888),
116. Cf. Abel, Orphica, p. 294, vs. 41 ff.; Kuhnert in Pauly-Wissowa. Damnameneus
occurs often in the papyrus; see below.
'0 See, for example, the interpretations ascribed to Androcydes by Clement,
l.c., and Hesychius, I.c., based probably on fancied etymologies; and Stickel,
De ephesiis litteris linguae Semitarum vindicandae commentatio, Jena, Universitatsprogram,
1860, a magnificent example of philological guesswork, in which the
six words are transformed into a Semitic epigram.
1 So apparently for Photius, Suidas, Macarius, and their successors.
Chester C. McCown
of Ephesus.12 To be sure, the tradition involved in the epithet
Cpocxrtos is not decisive. As Jessen pointed out, Artemis of
Ephesus is not by nature primarily a goddess of magic.13
Wiinsch suggested that ep4ca'a is to be derived from ELprJ.UinL
the sense of 'loosen,' rather than from "EoEoos.14 More
recently Professor Deissmann has explained the epithet as
derived from Babylonian epesu, 'to bewitch,' not an impossible
procedure in so international a science as magic. The term
e31ELa ypa/xuara, then, becomes a terminus technicus for
'magic formula,' while epesu, an unintelligible foreign word,
was associated with the name of the well-known cult center
and the traditional explanation grew up by a common etiological
process.15 Magicians and silversmiths might well
engrave incantamenta upon images of Ephesian Artemis,
sometimes upon one part, sometimes upon another, thus giving
rise to the varying and confused accounts of their supposed
place upon the great image in the temple. Professor Deissmann
does not suggest that the six words of the formula as
well as the epithet describing them came from Babylonia.
That would be a natural and almost necessary inference from
his ingenious suggestion. Until evidence is discovered, it
would appear that an Anatolian provenience is to be assumed.
For the Hellenistic age, certainly, the might of Ephesian
Artemis lay behind the ancient formula, but, more than that,
the power also of primitive magic and religion.
If the origin and meaning of the Ephesia grammata are
uncertain, their current value and use are clear. For the
literary sources they are powerful words of magic. Anaxilas
12 Gruppe, I.c., and Roscher, I.c. It is, perhaps, worthy of note that the lead
tablet given below, the earliest document that uses the formula, is from Crete,
also home of the Idaean Dactyli according to certain traditions; but that evidence
is too slight to decide the point.
13 Pauly-Wissowa, Realencyclopddie, v, 2761.
14 Rh. Mus. LX (1900), 84 f.
15 Deissmann, " Ephesia Grammata," in Abhandlungen zur semit. Religionskunde
u. Sprachwissenschaft Wolf Wilhelm Grafen von Baudissin zum 26.
Sept. 1917 iiberreicht von Freunden u. Schilern (Beiheft zur Z.A.T.W. xxxIII),
Giessen (1918), pp. 121-124.
130 [1923
The Ephesia Grammata
implies that they were written on amulets which formed a
characteristic part of the equipment of the rustic fop and glutton.
16 Menander says that they were repeated as aXet4apyuaKa,
'evil-averting spells,' by one who walked about those who were
being married.17 That is, they were spoken 'medicine.'
These two uses, then, as written amulets and spoken charms,
run through all the accounts that have come down to us.
According to Plutarch the 'magi' instructed those possessed
to repeat to themselves the magic words in order to drive
the demons out.18 Croesus is said to have saved himself
from the funeral pyre by naming them.19 As an amulet they
rendered the wearer invincible. At Olympia, so the legend
ran, an Ephesian repeatedly defeated a Milesian boxer until
it was discovered that the former was wearing on his ankle
his city's famous magical formula. When this was removed,
the Milesian easily won his three successive victories.20
Pausanias and Photius sum up their value correctly when
they say they were names or words having an "occult, evilaverting
sense," "a kind of occult power."21 "Those who
16 ev a-KvapioLs pa7rroio-at, opWv
'Eqpeo-ia ypal.LpaTKa aXa.
This is the end of a quotation given by Athenaeus, xII, 548c, from Anaxilas'
Lyropoeus, where the latter is poking fun at the rustic dude of his day. The
aKvrapLOF pawrToi were little " patches" of parchment upon which the " auspicious
Ephesian words " were written to " shield " the bearer. Yonge's translation
in Bohn's Library, Athenaeus, 877, is a beautiful example of modernizing,
as if the ancient fop carried a leather brief-case.
17
'E(piaLa roLs yaliovAcLv OrVTOS repL7raTre
Xeywv aXe~<cpaip/aKa.
Quoted by Suidas, s.v.; Meineke, Frag. Cor. Gr. iv, 181, ii; Kock, Cor. Att.
Frag. in (Leipzig, 1888), 108. Perhaps one should emend rots -ya JoimrLv to
batfovowatv with H. van Herwerden, Collectanea critica, epicritica, exegetica sive
addenda ad Theodori Kochii opus: Com. Att. Frag. (Leiden, 1903), 371, who mgkes
the suggestion doubtfully, referring to Plutarch, Mor. 85 B. Why not emend to
Tro0s ait.ootv, which is palaeographically, if not metrically, the easiest to
explain?
18 Mor. 706 E.
19 Suidas, Photius, Etym. Mag. s.v., Eustathius, I.c.
20 Photius, Suidas, Eustathius, locc. citt.
21 Eustathius, I.c., Photius, I.c.
Vol. liv] 131
Chester C. McCown
intoned them conquered in everything;" to wear them was to
carry a rabbit's foot, or, more classically, with Diogenianus,
"to bear a laurel wand."22 The one reference of the magic
papyri to the formula as such exhibits it purely as a "word of
power." In preparing a spell the magician is told, yp(a4e)
rbv X6(yov) Trv 6bpcaiKov a"(rTKKELC lcr TaUKEL ywv Ka'uX Ca3cpo lTrIotuv Xava
3\XE a,uLara KTX. Plainly aaKeLK ae TACLKELto, be read acKL
KaTaclKL, stands here for the whole formula, as paternoster does
for the prayer which it begins. To the copyist it had so little
meaning that he muddled the formula itself, even though it
was so common that a mere allusion was sufficient.23 Here
the spoken formula adds its magic power to another which
is to be written.
For the literary tradition, then, the Ephesia grammata are
words of power, and there is no evidence of their hypostatization.
When one turns to popular usage, on the contrary,
he finds them transformed from a magic formula, from mere
words, into active and powerful beings whose characters and
performances are known. An exquisite example of the employment
of these baiuoves as a written and spoken defense
against evil is found in the Cretan tablet already mentioned,
which dates from the fourth century B.C. The text may be
restored as follows:24
A(06a)Xlav ava ya(v valoVTa 66buo)v6E KEXEVW
(pEVy?ElI[EpV jE?TEPWOV O'LKWPCOVO 13LKCVaKCa pOXa-
Zy7Pr. T' lXELKCKaKOV Kal 'HpaKcXa 7rToXlropOo[v
'IaTpovP KaXco Kal NLKvY KaL 'ATro6XXwo[va
22 Diogenianus, iv, 77; Apostolius, vIII, 17.
23 Pap. Brit. Mus. 121, col. 13; Kenyon, Cat. of Greek Papyri in the British
Museum, I, 89, 11. 459 ff. Wessely, " Neue griechische Zauberpapyri," in Denkschriften
d. kaiserl. Akad. d. Wiss. zu Wien, Phil.-hist. Classe, XLII (1892), p. 35,
transcribes the passage correctly, but in his index (p. 81, aarKEL KaL rao'KeL TOV
'Oppa'LK6v Xo'yov) gives the impression that dOaKEKL al TraKeL is the X6yos 6p(poaiK6s,
which is obviously not the case. The allusion establishes the traditional order
of the words, as against Roscher; see above, p. 129, note 7.
24 The tablet was first published by Ziebarth, Nachrichten der Gesellschaft
der Wissenschaften zu Gottingen, 1899, pp. 129-133. It was reprinted with
fuller restoration and commentary by Wtinsch, Rh. Mus. LV (1900), 73-85.
In but a few places have I ventured to depart from the readings of the latter.
132 [1923
Vol. liv] The Ephesia Grammata 133
5 aLer', JcO' eXKet Terpa'yoS [AtI 'l aXov al'ya.
"Eracpos, "ETrapos, "Eirapos, (pevy', a,ua opev'ye, XvKaLva'
(E'vyE, KVwv, aua oa' ] , Kalt lp6K[X]oTros, arTe VyVOKOS
IUaLcvLiYEOb[vpo]Lav Trwv 7Irpos ajLiara avro'[] c'KaoTOs.
apKo[E] JE,lu r6gia[ECL,v aK]eTCr KVyE, "AOKL KarCa'KL'
10 ["AoKL KacTa' KL, [A'lota ALt, XlX]aolacvE ai0oX-y?
aloya iLa K K7lTro[cvX] av[eEre. [r]o[i] o/o a TeTpay
[K'KXfr atl- ool 6' oe[vjvo/a Tpianl,t eta veLS, Alo aKTj.
oX/3oOS C KC[a]Tcal5cia()a <e> 5taOl KCa [d]LatLrorv [E]KcoL
SppcE7[Eo]Xv7TO(5s ' OS) `X?p iuiaa KaiaapiLwrvo v KaT avsav..
15 < pat 7TETp a' TEcpayos>
Aa/Iuvay.evev, 6,uaMov oE KtaKCOaSe KOvTas aSvaycKa,
OS KA /LE air?7TaCLLK,a i Ot KLKa KOXXO3a\XQV-T
< EpaCKOTT[EpTOIV ] rEXELOTE7rTXOI xalpaCS atLiaCITOP' XAE(KEpaCS>
<XE[']orTOS O`V XEo8pcLKovros yXcoOEja y7vEo[v >
20 o[] /.uE KaTaXpu[ro'] rlXjrfer', ol Oi' fT' irE a]vlKrT,
OVTE lraTro? (pver?y vrayCo Pr alrpr vTOpE[tr] Twcv.
The readings are those of Winsch unless otherwise stated. ( ) indicates
restorations of lacunae; [ ] indicates additions, alterations, and restorations
of copyist's omissions; < >at indicates intrusions by error of copyist. WWunsch,
M-McCown.
1 AIO . AIAN. 2 OIKQIAP. 5 AIAIEFrAEAKEITETPArO2IITrTTTAIrArAAI2:
ater', wj' eXKec TeTpa-yos W, Aid claXov alya M. 7 IIPOKPOIPOc.
9 ETQIKTNE. 10 KATA2KIAA2IANENAA2IANENAMOAr-
Q2IAI. 12 KeKXr)TaMt (following the usage of the Test. Sol.), v'ierepov W.
13 KATAAMA,ITONQKOI. 14 MAKAPQN bis. 17 ol W. KOAAOBAAOI2I:
KoXXo0/3aXoLoTrL K oXX\\oaXooaWL 20 ol W OTTEEIIHNIKTQI. OuVT
M. 21 OTTEIIATh bis: V-YeT7rT-e`rayw W. 2INTOPAIIANTQNA: otivrop
ravTurcv M, aivropa 7r. W.
This is a real rz7rwc67a, carmen, an incantation composed in
the dactylic hexameter that was attributed to the Idaean
Dactyli and containing the Ephesia grammata which were
also ascribed to their ingenuity. I make bold to render the
rough and rugged lines freely:
Hear ye, malicious pack that infesteth the Aethalian country,
Forth from our homes, I say, away to your own dwelling places.
Zeus, the averter of evil, and Heracles, sacker of cities,
Healer, I call upon thee, on Victory, and on Apollo.
Chester C. McCown
5 Hear ye; the Lix of Tetrag the leaping flock now is bringing.
Epaphos, Epaphos, Epaphos, flee; thou she-wolf, flee also;
Dog, and thou Thieving Demon, thy fellow insatiate, together
Flee ye away, infuriate raging, each to his dwelling.
Keep off from the banquet the two greedy hell-hounds, O Aski
Kataski.
10 Aski Kataski, Aisia Lix, at milking time firmly
Drive ye the wooly flock home from the meadow. Thy name
is Tetrag,
But thou, 0 Wind, hast the name of Swift, a guerdon of Zeus.
Happily he who knows binding magic may pass down the
highway,
Shorn of his senses the silent, traversing the spirit-thronged
highway.
DD...aa..mm...n.n.a.a.m..m..e.en..n.e.u.e..,u.5.,. .d...o.. .tt..hh..oo..u.u. . ..tt.aa..mm...e.e. . .dbboyy ffoorrccee
tthhee wwiicckkeeddlyly ssttuubbbboorrnn, ,
Whoso may harm me and those who some charm would cast
o'er me to bind me;
................... ................... ..................

20 Whoso with ointments of magic would hurt me, to him be no

refuge
By ways whether trodden or trackless: to Earth, the All-spoiler,
I doom him.
The manner in which the tablet was folded as well as its
content shows that it was intended to be worn as an amulet.
The lack of connection in the text indicates that it was made
up of various charms compounded like a magic potion into
a sort of panacea for all the evils that beast, man, or devil
might attempt to work against the wearer.25 Lines 13 and
14 state most succinctly the importance of knowing and being
able to repeat such powerful names as the Ephesia grammata.
They express exactly the feelings of those who, according to
Plutarch, learned the names of the Idaean Dactyli and used
25 Line15, I take it, is the result of the copyist's effort to be certain that he
had at least once in his tablet the right form of the name of the wind spirit who
plays so important a part in the charm. Lines 18 and 19 are part of an a&XetPaplAaKovw
hich the maker of the amulet introduces as a sample of a KaTaxpLrTov or
allows to creep in by mistake as he unintelligently copies from his book of
recipes.
134 [1923
Vol. liv] The Ephesia Grammata 135
them as charms to drive away fear by confidently repeating
them one by one.26
The restoration of the corrupted lines of the tablet presents
in detail difficult problems, but for the present study it is
sufficient to note in general the treatment accorded the Ephesia
grammata. All of them are mentioned and, with the excep tion
of the proleptic Terpayos Atl and of A''oa, which is perhaps
sensed as an adjective, alaoa, in the order observed by Clement
and Hesychius.27 The restoration of line 5 is too uncertain
to be used in the argument.28 Wiinsch's ingenious restoration
of line 10, transposing the word Ati, which has lost its proper
place as the meter proves, may not be right in detail. But
it uses only elements already present in the tablet which, with
lines 11, 12, and 16, are sufficient for the purposes of this
argument.
Aski Kataski appears as an individual, since it is addressed
in the singular. Lix, or if Wiinsch's restoration is correct,
Aisia Lix, is a wind sometimes called also Tetrag. Tetrag
and Damnameneu are the two clearly defined characters.
Three points not affected by uncertainties of text are important
for the thesis here maintained: (1) the Ephesia grammata are
regarded as persons, for they are directly addressed, (2) they
are appealed to as beneficent, protecting spirits, and (3) Tetrag,
26 Mor. 85 B. The rationalizing explanation of Marcus Aurelius, 'Ectwcov
'ypa/LiUaort 7rapa-yyeXCLa EKELrTOo6 vvEX,v botro/Lu/6OKeacOa rTiV 7raXaCLLc TLOS rTv
aperT Xp1oaaarevwv(MY edit. xi, 26), fits closely enough this popular usage to allay
fear and might be read into the repetition to the newly married mentioned by
Menander. In the same way the Palestinian fellah, when he has to go out at
night where he thinks demons may be, " murmult beim Gehen bestandig Beschw6rungen
vor sich hin, betet ununterbrochen oder ruft vortwahrend Gott
oder einen Heiligen an."-Canaan, Aberglaube u. Volksmedizin im Lande der
Bibel (Abhandlungen d. hamburg. Kolonialinstituts, xx), Hamburg (1914),
p. 8.
27 The fact that aalia is an adjective, ' auspicious,' may explain this irregularity
and also the fact that it alone of the six words does not appear in the
magic papyri.
28 I have restored on the assumption that AI~ at the end of the line is for
Air, which is regarded as belonging to Terpayos, a genitive, the return of the
flocks marking the beginning of the night and the consequent need of protection
against the demonic powers that were active in darkness.
10
Chester C. McCouwn
or Lix, is a wind.29 The bearing of these self-evident conclusions
will be seen when we turn to another document dating
some six or seven hundred years later.
The Testament of Solomon, a Christian combination of
Jewish and heathen demonology and magic, probably belongs
to the third or fourth century after Christ.30 It is a thesaurus
of magico-medical, demonological, and astrological lore,
strung together on the slender thread of a narrative as to how
Solomon, by means of his God-given ring, secured power over
the demons and knowledge of their characteristics, learned the
manner in which each was to be circumvented and made
powerless, in other words, cured, and then used them, by way
of punishment, in the building of the Temple. It comes to
us in manuscripts of the fifteenth and sixteenth centuries,
six of which contain a passage which I read as a description of
Lix Tetrax, the wind demon specially singled out for attention
in the Cretan tablet. The chapter which follows is a fair
specimen of the whole:
VII. Kal e?VXo'y'o(as TrO e0Po v eyCo OXO/.tUPEvK XEXvuc7ar apceval
gOL ETEpov aai,LOva'- KCaL?LX OE WTp TpocaTrov 1LLOV. Kal v7PT O TrpoaUrO
ErtL(EpWPV TC( aCpL Vt/77XOv KaLL TO bV7rXELTOV TOV OCLarTos EiXOV/IEov
cbOel KoXXias. 2. Kal Epprl7E oTpaCLTcoras OVK OX\iyov Kal 7ETepE
KaL Xca3pov KOVLOpTOpV cLaO Tr3S yjSy KaC avefpepv avoW KCaL 7roXXa
eppLT7r'eV& 7rl To '!e' Oafi#eloOat' Kal ed7rov rtiva eXo E&wpUwsr o~taat;
Etr 7roXV. 3. KaLL alvaaY'TaLTOS 1.OV eTvaaCL xa/taL KaT' EKELVOV TOV
TOTrOV KaLL EcppdayLOa T& ) aKTvXL'LO, TfOV Eov, KCaL OVTWS EaTr 7]
avpa cKELPvr. TOTrE 7pwTrrj a aLVTO XeyWv Vj TrLS el; KaLL OVTOS
KOVLOpTOV TLvClatS alTEKpirl /IOL- TL OXetLS, a3CoXELV ZoXoJ ov; 4.
a7reKpl67v 6 aauc LT eirE .totL XEy' eL KtayCo aYe PWTcorv Oe\Xc. OVTircS 6
EVXapLaPTjr TO Oej T? aooplcraPr/V Tie rppOs Tal f3ovXas aCvTPv aTLOKpL-
29 Following the usage of Clement and Hesychius and, as I believe, of the
Testament of Solomon (see below), I assume that AIt and TeTpay may appear
together, as does "AOKLK aractK, Ait being a secondary, Terpa&yth e important
name for the one character. Whether Aski Kataski is to be taken as one name or
two is indeterminable. In the literary tradition it is two.
30 See discussion of the date in my edition, The Testament of Solomon (Untersuchungen
zum N. T., herausg. v. H. Windisch, No. 9), Leipzig: Hinrichs,
1922, pp. 105-108.
136 [1923
Vol. liv] The Ephesia Grammata 137
vaacOa. 'e(prO6sl io 6 aloatuov' 'yc KaXovcatLA iT Terpa6. 5. elrov
6e abvT' rTs r 7rpals -7ov; C'pcr7 dpcaz' V rovs (TKOpT7rl Kal aTpo'6ovS
7rotl KaL 7rUp a7rTco KaL a'ypovs ,zrwvplw Kat OLKOVSK arap-y(. irt
TrXeLaTOV b5 EXCO Tr'v rpCdLt EE v OpeL. edav 6 Katpov evpoW, viroSv'wx
ls ycowvlasT roXCovV VKTCKa al 7r,JEpatv'r el7tly 'ap Yovozse yatXov7.T OV
6. E7rov &c avTTrw' v rolWt a-TpcW KELZoLa; 6 / EL7rT' ELS avrT To
aKpov TOV KEpaTOs T7S a\EX'vr/s TO kv Tmr VO'TC EVpLaKO1AEPvEovK ELI tOV
ro &a-Tpov. OLOT6La o(pra Xara Tro r-jcqLTpLTalov rpoo-erTaXOrv
avlc/iiOatl. Lta TOVTO ltb6vTES roXXol TCoV avOpWCirCwv exovTat elS Tov
7l7LrTpLTaloi Ebv roZs rTpalv o6v6oaaI TOVTOLS' ovXrTaXa' 0aXXaX'
EcEXXaX' Kal lcjlta, avrovs. 7. eL7rov oE avTCr? EycJ IoXoaoiv' oTre
OVvO \XELKS aKOVp'yelVE, v rLvi KaTapYEloc'a OL6; f fp1' ev Tw apX'ayy'Xw
wc Kal o 7j7tLTpLTaCos Traverat. T7r7ppcOrtlCa 5 avrovL' 7vr olv

ovo6iarT KaTapyy'catc; 6 o elreV' E-?v T vo6Yar Tov apXa'yypeXov

'A3'a"X. 8. Kal EW7reappatyLaa TOv Saiaova Kal EKEXEvaa aVrTO
Xltovs apTra'ltEL Kal ElS T7a vltrX7]\a TOy vaov aKOv1IleLV TOiS TEXvlTaLS'
Kal ava'yKa6tEo'evov To 6aL,uo'vov Ta 7rpoo'ETrae'y,e a avTrc T9rolte.31
There are many more or less important variae lectiones in the
six manuscripts, but the only point of significance for our
present problem is the name of the demon here described. I
have "restored" Al~ TeTp6a, first because I believe the evidence
of the manuscripts makes it the probable reading, and
second because there are important resemblances between the
demon of the Testament and the Lix Tetrax of the tablet.
First, do the manuscript readings justify the emendation?
The fifteenth century copy of the Testament in the library of
the Earl of Leicester at Holkham Hall 32 gives the name as
77 Xt TErppas. Harleian Ms. 5596 in the British Museum, also
of the fifteenth century, reads els Xe' reqppLaaOat, and, omitting
the first nine words of the next section, takes the infinitive
with what follows (avOpWcrovsK, TX.). Two other manuscripts
of the same century, both copied from the same exemplar in
Italy, one now No. 3632 in the library of the University of
3 For the manuscript variants see the edition cited above, pp. 28 * ff.
32 Since the publication a letter from the librarian, C. W. James, Esq., informs
me that the manuscript came from Venice to " Mr. Coke of Norfolk"
in 1721, having formerly been the property of a Cretan priest, Mark of Morzenos.
Chester C. McCown
Bologna, the other No. 2419 of the "Anciens Fonds Grecs"
of the Bibliotheque Nationale in Paris, read 6Xl Terspas. Two
manuscripts, one in Jerusalem, Sancti Saba No. 422, the other
No. 38 of the "Anciens Fonds Grecs," read e-yce LdurtT Ow 7veva Trjs
rTqcpas, instead of -ycJ KaXo/uaLt Alt TerpaZ. Palaeographically
there can be no objection to my conjecture. The
resemblance is so close as to make the change easy from unintelligible
TcrpaT to familiar reppas, 'ashes,' especially as the
dust of the windstorm might easily suggest the latter. As to
the first name, in such late and carelessly copied manuscripts
as these, one need not seek a full explanation of such small
alterations. It is entirely in keeping with the general character
of the manuscripts that the first, third, and fourth should
preserve an unintelligible reading which closely resembles
thd original, that the second should completely confuse it by
pure carelessness or ineptitude, and that the last two should
alter it into something reasonable and intelligible, but quite
other than the author intended.
The second question is, Does the character of the demon in
these two widely separated documents, the Cretan tablet and
the Christian Testament, justify the identification? There
are several reasons for answering in the affirmative. First,
although the literary references to the Ephesia grammata
imply that they are words only, our two non-literary documents
plainly regard them as powerful demons, or spirits, each with
his own individuality. In both Lix Tetrax is a wind and, most
significantly, the "gift of Zeus" of the tablet is in the Testament
"offspring of the Great One." That the beneficent gift
of Zeus, the cool evening breeze that with its eddying gusts
brings the flocks home in safety, becomes in the Testament an
evil, whirlwind demon that burns and causes fever is entirely
in keeping with Jewish and Christian demonology, which
turned the spirits and gods of Heathendom into devils. The
notable fact is that, in the Testament, contrary to all its usage,
when men pray to the semitertian fever with certain magic
words, Lix Tetrax cures them. With the passing of the cen-
138 [1923
The Ephesia Grammata
turies and the coming of the new religion, Lix Tetrag has
undergone alterations, but the storm-demon of the Christian
Testament is the lineal descendant of the Cretan ailuwv of the
evening breeze.33
The numerous occurrences of the word Aa,uva,uevei in the
magic papyri and amulets prove the spirit's popularity. He
was probably thought of as the 'Tamer,' his name being
derived from SauLvacw-ba,ui&aaso ,i n the Cretan tablet. In this
case literary and popular tradition agree, for he was one of
the artisan YyoErcs 'IbaZot, later one of the three Idaean Dactyli
whose names are handed down as,
KEXAtLAs aluvaaL?evv's- re Iuelyas Kalt birpf3Los "AK,wLV,
and also one of the Telchines.34 Nonnus relates a love affair
in which he played the hero.35 Like other alaitovrs, in the
magic papyri he is little more than a name, but fully as real a
person as the great deities. In Pap. Leiden. v he is invoked
along with Iao, Sabaoth, Adonai, Chettaeus, Axiothof, the
seven vowels, and others, as "greatest god,"36 in the Paris
Magic Papyrus immediately following Adonai and Zeus.37
One can have little doubt, therefore, that for the non-
Christians of the fourth century A.D., as for the Cretans of
the fourth century B.C., the Ephesia grammata were still
living and active spirits, as powerful and beneficent as in the
33 The word X1l occurs at the beginning of the Paris Magic Papyrus (1. 3)
but in so uncertain a connection as to be unintelligible.
34 Scholiast to Apollonius Rhodius, I, 1129 (Merkel-Keil, 1854, p. 371),
Lobeck, Aglaophamus (1829), p. 1157. The chief sources are Diodorus Siculus,
v, 6 f., and Strabo, x, 3, 22. See Lobeck, op. cit., pp. 1156-1181, Kern in Pauly-
Wissowa, iv, 2018 ff. and Triimpel, ib. iv, 2058, Gruppe, Griech. Mythol. u.
Religionsgesch. 455, n. 1, 1522, n. 4.
35D ion. xxxIIi, 326. Perhaps for this reason he appears in the a&yo,-yio, r
love charm, Paris Mag. Pap. 2772, 2779.
36 Col. 19, ed. Leemans, p. 31.
37 Lines 2772 f., Wessely, Griech. Zauberpap., 114. A few of the many
references to Damnameneus are listed by Wessely, op. cit., Index, s.v. A
figure almost identical with that in Kopp, Palaeog. Crit. iv, 203, is found among
the Intailles Gnostiques, in the D6pt. de Mbdailles of the Bibliothbque Nationale,
Paris, No. 2181.
Vol. liv] 139
Chester C. McCown
time of their prehistoric origin. Koch may have been right
in saying that they were the names of the Idaean Dactyli,38
but in any case they were mythological beings, probably pre-
Greek deities. Putting together the usage of the Cretan tablet
and the magic papyri, popular documents which may be expected
best to retain primitive forms, one concludes that they
were called caKL, KaTraOKL, CauLa, Xi, reT7pa, a,UJva/uLeve. The
form of the words together with the literary tradition suggests
that they are of Indo-germanic, probably Phrygian origin.39
They were in some way associated with the great goddess of
Ephesus, were gradually Hellenized in form and degraded in
place, while their original meaning was forgotten until they
became for the educated a mere formula and for the Christians
evil spirits, with just a touch of their former benevolent character.
Perhaps the excavations which we hope may soon
become possible in rejuvenated Anatolia will eventually recover
them, or possibly they may be found when the tantalizing
Cretan script is deciphered.
38 Cor. Att. Frag. ii, 1, 268.
39 Gruppe, Griech. Mythol. II, 884, n. 2, Roscher, Philologus, LX, 101, " Nachtrige
" to p. 89. The latter notes resemblances between them and certain
Phrygian names, but thinks the words may come from a non-Greek, pre-
Phrygian race.

140 [1923